Help with reripping Matrix: Path of Neo by Ultrafighter at 3:26 PM EDT on April 18, 2018
Hi guys, I'd like to get some assistance with this title and I believe that everyone will benefit from addition of better rip to VGM archives. My 1st complaint about this set is absence of quite a few segments for interactive BGMs, simple removal of duplicates from this release resulted in numerous unexpected gaps in tracks.
Also original ripper, Alpha23, reinterleaved given headerless audio with interleave that equaled 0x6000 to files with custom SS2 headers & 0x10 interleave; now after both TXTH & TXTP have been introduced I guess there's all more reason for a proper and final rerip.

Just as Alpha usually did with his rips he left many scripts + utilities that he used for ripping in the final package and they can actually be found here (copied from !admin.7z contained in aforementioned mirror of his gamerip). A few BMS seem to be dedicated to PC/Xbox ports of the game, matrixext.exe is possibly unused too & there's no reinterleave.bms but this is what we've got so we have to work with those helpers / leftovers if we want to devise some fresh method of ripping VGM.

My reconstruction of Alpha's actions has usage of WAD_split_PS2.bms on all the *.wad bigfiles in the game (I've only tested it on the ones containing dynamic BGM though) set as the 1st phase. Resulting *.ss2 are already playable but unfortunately there's a pretty noticeable gap (tiny section of silence) in the end of most segments.
Then gen_split.bms was probably used to divide those SS2 tracks by stop marker but this is where his procedure stops being very clear. There're always 2 such markers closer to stream's end but we only have to get rid of everything behind the 2nd one, could he set that BMS differently or simply stitch 2 output files without extensions together (the 1st file should have "SShd" signature in the beginning & that stop marker in the end, the 2nd one should have zeroes at file start & the same marker at end, the 3rd one has to hold zeroes only and should be quite small)?
Well let's assume he's got those slightly diminished SS2s (maybe he created some way of automatic joining of parts #1 & #2 into a single file, maybe he utilized something instead of gen_split.bms). And finally for such trimmed versions of interactive music segments (this is what the term "IMS" supposedly means in both Enter the Matrix & Path of Neo) he had to tweak header a bit because file size is now incorrect. There's some bin_conv.bms in his !admin.7z which might have been designed for this task but I didn't get good results with it; however, Alpha23's append_header_SS2.bms + func_header_SS2.bms from XeNTaX seem to do the job fine.

So can you please help me out? Some brand new BMS for extraction of headerless sounds from those WADs would be the best solution I presume, maybe it'll even avoid writing unwanted sections of audio archives to disc thus nullifying all the aforementioned trouble on steps #2 & #3...
But there's additionally NSGLOB.WAD with multiple non-interactive BGMs, all the tracks are stereo too but frequency varies a lot (5 unique values, from 44100 Hz to 20 KHz, in 83 rather long SS2 files) so these *.wad archives might really have some sort of headers for contained streams. What do you think?

Bye and good luck!

PS. Samples (2 full WADs).

edited 3:34 PM EDT April 18, 2018
by Ultrafighter at 8:23 AM EDT on June 6, 2018
Hi all! So to sum it all up: the problem's been solved & I've found 2 or even 3 ways of ripping this title and will list them right now then post my complete rip (done using the very 1st method as to preserve original headerless data by any means possible). I'll also post the newest ripping kit to automate all the steps required to rip soundtrack from the game.

These are 3 ways to extract individual sounds, BGMs & dynamic music segments out of given WADs from PS2 version then make them playable:
A. MatrixExt & VGMstream + TXTH:
1) copy all the initial *.wad bigfiles to HDD because matrixext.exe by Snakemeat won't work otherwise;
2) unpack them with MatrixExt.Exe:
a) for IMS\B#__~1.WAD & NSF_EN~1.WAD: simply use MatrixExt.Exe* (look for A1.bat in my ripping kit);
b) for NSGLOB~1.WAD: first split it with Snakebite.exe by Snakemeat using A2.bat then use MatrixExt.Exe once more* (just run A1.bat again);
* it's needless to say that an executable, a batch file & the files you want to process in batch-mode should be in the same folder for my BATs to work properly;
3) play resulting BINs with my *.txth file via VGMstream plugin for FB2K / XMPlay / Winamp:
a) to do that merely move .BIN.txth to the folder containing *.bin files then add BINs to playlist of your player;
b) A few smaller voice clips or movie tracks might sound weird with .BIN.txth so that you're welcome to experiment with interleave value (change extension of troublesome sounds to *.bi then try changing value in the line #2 in .BI.txth, save this text file then add (probably) fixed BI file to playlist for testing purposes).

B. WAD_split_PS2.BMS & MatrixExt:
1) extract all the WADs from IMS folder (B#__~1.WAD ones) & NSF_EN~1.WAD with WAD_split_PS2.bms by Alpha23 (using QuickBMS.exe obviously, you'd better use B1.bat for batch-processing purposes) then play resulting SS2 files with VGMstream (simply add them to playlist as no TXTHs are needed in this case);
2) remaining audio:
a) split NSGLOB~1.WAD with Snakebite.exe by Snakemeat using B2.bat;
b) use WAD_split_PS2.bms on each of 6 smaller WADs made by Snakebite.exe (you can use QuickBMS GUI to process only these newly created files but I'd advise you to simply run B1.bat once more) and use VGMstream for playback of resulting SS2 files (merely adding them to playlist should do the trick just fine).

C. MatrixExt + bin_conv.BMS:
1) copy all the initial *.wad bigfiles to HDD because matrixext.exe by Snakemeat won't work otherwise;
2) unpack them with MatrixExt.Exe:
a) for B#__~1.WAD & NSF_EN~1.WAD: simply use MatrixExt.Exe (look for C1.bat in my ripping kit) to get multiple BINs out of each WAD;
b) split NSGLOB~1.WAD with Snakebite.exe by Snakemeat using C2.bat then use MatrixExt.Exe on each of 6 smaller WADs made by Snakebite.exe (run C3.bat);
3) process all the *.bin streams with bin_conv.bms by Alpha23 (run C4.bat) then play resulting SS2 files with VGMstream.

You can get my gamerip from here and ripping kit from there.
Cheers!
by Anterag at 4:01 PM EDT on June 7, 2018
AWESOME rip! Keep up the good work! Love the game soundtrack
One question though: are the interactive music system segments (IMS) in order ?
Thank you in advance.
by Ultrafighter at 5:40 AM EDT on June 8, 2018
Well I left them all in the same order in which they were extracted by MatrixExt.Exe and apparently in the very same order in which they're initially stored in WADs.
If you meant that kind of order which provides the best listening experience then I guess not: if you load all the segments for a select level in your player and play them from the 1st to last you'll quickly notice that there's typically 1 or 2 variants of intro in succession, then a dozen of segments serving as main part of a track, then some longer segment with a fadeout (or maybe 2-3 supposed outros in sequence), then another succession of segments which can be considered continuation of that previous theme, then an outro followed by an intro for a completely different composition and so on. I think it means that you'll have to invest quite some time in your rip to make it look and sound like completely tagged gamerip by Ggctuk2005 of FFS forums.

Anyway when we're dealing with such interactive BGM systems I believe we can't really complain about lack of proper segment / track order because there's no such thing as strict order of those short segments with fixed length, outros and intros; that's just the way it is I presume.
Later!

PS. I'd advise you to use XMPlay & simply add entire folder with IMS segments to playlist, I'm not sure that FB2K or Winamp are going to keep the right order of those sound snippets. You see numbering is hexadecimal and starts at 0, not 1 and correct order should be like this:

B7____~1_00000098.BIN
B7____~1_00000099.BIN
B7____~1_0000009A.BIN
B7____~1_0000009B.BIN
B7____~1_0000009C.BIN
B7____~1_0000009D.BIN
B7____~1_0000009E.BIN
B7____~1_0000009F.BIN
B7____~1_000000A0.BIN
B7____~1_000000A1.BIN
B7____~1_000000A2.BIN

then

B7____~1_000000A8.BIN
B7____~1_000000A9.BIN
B7____~1_000000AA.BIN
B7____~1_000000AB.BIN

and so on.

I've been using XMPlay for years now and can confirm that it always keeps proper order of numbered files even if numbering is hexadecimal.
That's all for now, bye!

edited 5:57 AM EDT June 8, 2018
by marcoxD95 at 4:04 PM EDT on August 4, 2018
Would it be possible to reupload the soundtrack?
MEGA sadly has the habit to delete everything very fast.

I ever since looked for a proper rip of this soundtrack! It is one of my favorite videogame soundtracks!
by datschge at 5:43 PM EDT on August 4, 2018
@marcosD95 - Ultrafighter's rerip is already on joshw.info, search for it on vgm.hcs64.com
by Ultrafighter at 3:22 AM EDT on August 5, 2018
Yep it was me who deleted it as soon as I saw it appear in JoshW collection (and to clear up space too). It's here now.
Bye!

PS. I'd advise you to move all the *.BIN files from IMS/Duplicates/ folder to IMS/ before you start listening, maybe I should have left all the interactive segments (even dupes) in one directory after all.

edited 3:28 AM EDT August 5, 2018
by marcoxD95 at 1:51 PM EDT on August 5, 2018
Thank you so much Ultrafighter!

Was able to acquire the rip and extract the SS2 files with your ripping tool (C4.bat). Everything seems to be playing just fine.

I am actually astonished. I remember working with this soundtrack many many years ago and went total crazy due to the absolutely annoying codec and the flawed pon extractor not being perfect at extracting the sound. I think there was sometimes some white noise or something like that or the sound not being extracted at lossless quality.

I didn't notice anything of that sort with that rip yet. Also when converting the SS2 files into Wav/Lossless I actually get proper lossless audio files with proper wavelenghts and hights! Already viewed the audio stream in a spectrum analyser tool. This is astonishing.

It must have been a lot of work to get it working properly like that.

Now the only problem left is that someone would have to actually merge the music properly into tracks without any incorrect transition segments or other sound effects in it and THIS is probably an giant sized project considering there are like 2500 files or something like that. I would theoretically be up for this but I didn't play the game in ages so it would mostly be guesswork and not reflecting the real music. Unless I take previous rips as guiding help and try to maybe add some missing segments into it. Either way it's an absolutely crazy amount of work.

I didn't get the part about the stop marker and those transition infos inside your notes info. You mean that the segments are not always in the correct order?

Anyway, this is probably the best rip for the game yet and its near perfection, just missing the proper transitions and merged tracks. Maybe I will try working on it but don't think I ever finish it properly. Dunno, it's just too much.

Wish someone who is already skilled in this or perhaps even released a rip for this game could redo it with all those files here, even lossless. I remember 2 major rips for this game in the ffshrine forum. They were both really good but used this "meh" pon extractor and the audio quality wasn't perfect.
by Ultrafighter at 2:16 PM EDT on August 6, 2018
Do you mean you got my rip with numerous BIN files & a few TXTH ones then converted the former to SS2 using C4.bat from my ripping kit and played resulting tracks in VGMstream? That wasn't necessary since those BINs are readily playable via that plugin but whatever, if it suits you it's fine by me.

I added quite a bit of text about stop markers to that "! Notes.txt" doc because those BINs don't always have fluent transitions between segments when I play back sequence of those segments in a player with VGMstream installed. Is it just me or someone else hears a split second silence between B7____~1_0000009A.BIN & B7____~1_0000009B.BIN, then B7____~1_0000009C.BIN & B7____~1_0000009D.BIN, etc.? This kind of obstacle really prevents tracks from fluently flowing into each other.
I tried manually deleting those markers from file end and re-save those trimmed audio chunks (for my personal use only, the set I submitted only includes original unedited streams extracted with MatrixExt.Exe from initial WAD bigfiles) but this doesn't always help, also it's obviously quite bothersome and time-consuming to edit dozens of segments this way (by hand I mean).
So I'd say that only lack of perfect transitions doesn't allow me to call rerip of mine the best & the ultimate offering, as I said Ggctuk2005's version has merged compositions & detailed tags (I still wonder how he found out who did each track) but my ver. should provide the best 1:1 preserved quality of sound and something missing from earlier Alpha23's rip (movie tracks and a bit more). Take your pick.

So long!


Go to Page 0

Search this thread

Show all threads

Reply to this thread:

User Name Tags:

bold: [b]bold[/b]
italics: [i]italics[/i]
emphasis: [em]emphasis[/em]
underline: [u]underline[/u]
small: [small]small[/small]
Link: [url=http://www.google.com]Link[/url]

[img=https://www.hcs64.com/images/mm1.png]
Password
Subject
Message

HCS Forum Index
Halley's Comet Software
forum source
Generated in 0.0029s;